Learning Adversarially Robust Representations via Worst-Case Mutual Information Maximization.

Sicheng Zhu, Xiao Zhang, David Evans

Browse the full ICML paper archive.